High-Probability Distractor Patterns
Key Takeaways
- Most strong distractors are partly true but fail because they skip a step - assessment, valid data, ethics, or implementation verification.
- Reject answers that pick a procedure from topography, diagnosis, convenience, or stakeholder pressure rather than from function.
- Treat absolute wording (always, never, immediately, guarantee, most intensive) as a flag to verify the vignette actually justifies it.
- On 'best next step' items, an option can be a good LATER action and still be wrong NOW - sequence is the test.
- Better options tend to define, assess, verify integrity, consult, train, monitor, document, and modify based on data.
Distractors That Sound Right
The most dangerous distractors are fluent: they use correct terminology but apply it at the wrong moment. "Use extinction" can be correct - but usually only after function is known, a replacement behavior is being taught, risks (extinction burst) are managed, and implementers can follow the plan. Drop any of those and the same words become the wrong answer.
The exam frequently asks for the best next action. On those items, an option can name a perfectly good later step and still be wrong now because something earlier in the chain is unresolved. Sequence is the thing being tested.
A second source of fluent distractors is partial truth. The option states something that is correct as a general principle ('reinforcement is preferable to punishment,' 'token economies are evidence-based,' 'extinction reduces behavior') and attaches it to a case where it does not fit. The statement is true; the application is wrong. You cannot eliminate these by judging the sentence in isolation - you have to test it against the vignette's specific function, data, and constraints.
A Catalog of Recurring Wrong-Answer Patterns
Learn these by name. When you can label why an option is tempting, you can eliminate it fast.
| Pattern | Why it fails |
|---|---|
| Topography-only treatment | Same behavior form can have different functions; pick from function, not form. |
| Diagnosis-only decision | A diagnosis does not identify the maintaining contingencies. |
| Measurement convenience | The easiest data may be invalid for the actual decision. |
| Graph overreach | A trend in one phase is not a demonstrated functional relation. |
| Ethics bypass | Consent, assent, competence, rights, or safety are skipped. |
| Integrity neglect | Treatment is judged 'ineffective' before fidelity is verified. |
| Staff blame | Staff behavior is moralized instead of analyzed functionally. |
| Stakeholder appeasement | The procedure is chosen to satisfy a caregiver/funder, not the function. |
| Restrictiveness jump | A restrictive/punitive option is selected before least-restrictive ones. |
| Absolute claim | The option ignores context or legitimate exceptions. |
Red-Flag Wording
Certain words signal that an option may be skipping analysis. They are not automatically wrong, but they earn a second look:
- immediately implement / discontinue without notice - may skip consent, transition, or data.
- guarantee / always / never - behavior is probabilistic and context-bound.
- the most intensive/restrictive option - violates least-restrictive logic unless justified.
- collect no more data / ignore the data - removes the basis for decisions.
- share details (with an unauthorized party) - confidentiality breach.
When you see one, return to the vignette and ask: do the stated facts actually justify this? If the facts are missing, the word is doing the work, and the option is probably a distractor.
The inverse is also worth noting: hedged, process-oriented language ('assess,' 'monitor,' 'consult,' 'with consent and training') is not automatically correct, but it correlates with better answers because it respects the chain. Do not pick a hedged option blindly - it can still target the wrong function or skip the actual ethical hinge - but treat a stack of absolute red-flag words as a strong reason to look elsewhere first.
The 'Good Later, Wrong Now' Trap
The single most common integrated-item error is selecting a valid action at the wrong point in the sequence. Extinction, punishment, fading a prompt, thinning a schedule, or discharging a client can all be correct - eventually. The 'best next step' phrasing is asking which action the current facts justify.
Work an example. A vignette establishes function (escape), a taught replacement mand, and stable data, then asks the next step; an option says 'begin schedule thinning.' Reasonable later - but if the replacement is not yet fluent and reinforced densely, thinning now risks losing it. The stronger 'now' step keeps dense reinforcement until the mand is established. Train yourself to ask: what does the vignette say has already happened? The unfinished earlier link tells you which option is premature.
The same logic governs generalization and maintenance distractors. 'Program for generalization' is a fine later step, but if the skill is not yet acquired in the training setting, it is premature; you cannot generalize a response the learner cannot yet emit. And 'fade the prompt' is correct only once the prompted response is reliable. Each of these is a real procedure misfiled to the wrong moment - the recurring shape of the strongest distractors.
What a Stronger Option Looks Like
Better answers cluster around verbs that advance the case while protecting the client: define, assess, verify integrity, consult, train, monitor, document, modify based on data, and collaborate. They are specific enough to act yet cautious enough to preserve analytic confidence and welfare.
A reliable two-step elimination: (1) cross out any option that skips a chain link the vignette left open (function unknown, integrity unchecked, consent absent); (2) among the survivors, choose the least-restrictive, most function-based, data-anchored next step. This converts a four-option guess into a principled decision and protects you from the fluent distractor that 'sounds like a BCBA wrote it.'
A vignette states a learner is diagnosed with ASD and engages in hand-flapping; an option reads 'Because the learner has ASD, implement response interruption and redirection (RIRD) for the flapping.' Why is this a classic distractor?
On a 'best NEXT step' item, baseline shows three highly variable data points and the option 'begin the intervention now to save time' is offered. Why is it tempting but likely wrong?
Which option below shows the features of a STRONGER integrated-case answer?